Big data and apple pie
Correcting a story in a TED Talk
In my TED talk “Big data is better data” in 2014, I begin with a lovely anecdote about apple pie. It goes like this:
America’s favorite pie is? (Audience: Apple!) Of course it is. How do we know it? Because of data. You look at supermarket sales of 30-centimeter frozen pies and apple wins. No contest. But then supermarkets started selling smaller, 11-centimeter pies, and suddenly, apple fell to fourth or fifth place. Why? Think about it. When you buy a 30-centimeter pie, the whole family has to agree, and apple is everyone’s second favorite. (Laughter) But when you buy an individual, 11-centimeter pie, you can buy the one that you want. You can get your first choice. You have more data: you can see something that you couldn’t see when you only had smaller amounts of it.
Since then lots of people have contacted me: asking if apple isn’t the most popular, then which is? I’ve done a bit of digging and have a feeling the story isn’t quite what it seems.
Specifically, the anecdote comes from a beautiful appreciation of the late economist Walter Oi, who died in 2013 (here). Oi studied monopoly pricing and economic costs; his most notable contribution was to show why an all-volunteer army is more efficient than the draft, which helped end conscription in the US (described here).
The tribute by the University of Chicago (his alma mater) noted:
He was known for using simple, real-world cases — and sometimes seemingly irrelevant information — to understand pricing. For instance, he posited that the preference for apple pie depended on the size of the pie.
But one can find no references to any academic work on this.
However a clue to its veracity crops up in another wonderful appreciation of Oi, by his former colleague Steven Landsburg at the University of Rochester (here). In his words:
Walter loved facts so much that he sometimes invented new ones, because the world could always use more.
Prof. Landsburg gave an example of Oi’s prankster-ish nature, especially about economics.
One day he walked into the department coffee room and announced that “A one hundred pound man and a three hundred pound man have exactly the same quantity of blood.” When this was met with considerable skepticism, Walter responded as he always responded to skepticism — by repeating himself more forcefully. […] A couple of my more enterprising colleagues went home and checked their encyclopedias that night, and came back the next morning to report that according to authoritative sources, a man’s blood volume is roughly proportional to his body weight. Walter’s response: “Nope. A one hundred pound man and a three hundred pound man have EXACTLY the same quantity of blood.” If you watched carefully and didn’t blink, you might have caught him suppressing a smile.
He was an eccentric. The University of Chicago’s Economics Department has long had an annual “Walter Y. Oi Award” for the most irrelevant question (presumably honoring one of his charming foibles).
His puckish humor was perhaps a way of relishing a life tempered by seriousness: at 13 he was placed in a World War II Japanese internment camp in Colorado. He lost his eyesight in the 1960s.
Even if his story of pies is a bit like his exchange about blood, and invented for the sake of semi-scholarly fun, one story is true, which attests to his courage and values (described here). As a junior faculty member at Iowa State University—big farm country — he chastised colleagues for promoting agriculture as a growth industry when it was in decline…and he lost his job the subsequent year.
That’s the hallmark of an economist who is true to his principles — and to the data.